 Well, thank you so much for the invitation to come down and speak with you today and I apologize. I wasn't able to make it in in person. I hope to do that at some, some date in the future. But today, I'm going to talk about valvular heart disease, anticoagulation and pregnancy and this victim that I was taught in 11th grade of there ain't no such thing as a free lunch. So thanks. Thanks for joining today. So it helps to start by thinking about who were who were talking about. So we'll begin with a patient a hypothetical patient a 28 year old woman with a bicuspid aortic valve who comes in to see you as a new patient. Reporting that she played tennis in college and that really just goes for walks and thinks she might be slowing down a little bit. A normal left and trigger systolic function with an EF of 60% on her echocardiogram and a bicuspid aortic valve, which meets the criteria for severe aortic stenosis with a maximum velocity of 4.3 meters per second and a mean systolic pressure gradient of 42 millimeters mercury. She's recently married and as many of you can predict her first question to you is Doc, is it okay if I get pregnant. And this type of patient is probably familiar to a lot of us and it's not a rare patient if you work in the field of congenital heart disease or certainly cardio obstetrics or valvular heart disease or general cardiology even because there's been these enormous advances in the treatment of congenital heart disease for infants and children. And it's now the expectation that virtually all women born with congenital heart disease are going to survive until adulthood. And the vast majority of them are going to arrive at adulthood with really good health expecting normal life expectancy, living active lifestyles and aspiring to live a normal lifestyle and that includes having kids. So, increasingly, it's becoming common for us to have to feel this question to try to address the risk of pregnancy and women with structural heart disease. So just as a brief overview of what we'll be talking about today, we'll be talking about preconception evaluation of women with aortic stenosis. We'll be talking about valve choice and women considering pregnancy. We'll be talking about anti coagulation for mechanical heart valves and pregnancy. And then at the end we'll shift gears a little bit and we'll talk about the style in which we give counseling to patients, and how that affects their choices and this is not entirely specific to the questions of pregnancy or congenital heart disease but we'll just review some of the data about how, how the words we choose affects what patients here and what they decide. So coming back to our patient with severe aortic stenosis who wishes to get pregnant it's not that there is no choice there are no good options of what we might do for a patient like this is that there's unfortunately a lot of options, but each of them comes at a price. So, perhaps if our greatest concern is the safety of the patient we can anticipate that severe aortic stenosis will pose some risk. And if our number one priority is trying to shield our patients from risk. We could go ahead and discourage patients and tell them and tell her, you know, I don't, I don't think it's a good idea for you to become pregnant. The problem with that of course is your patient is unlikely to be happy with that it might be unnecessarily restrictive. And as we learn over and over again there's certainly no guarantee that our patients are going to listen to us if we give them advice. So you might end up with a unhappy patient or you might end up with a patient who goes off and becomes pregnant. It might be outside of the care that you might be able to provide. So if patient satisfaction is her number one priority and we all want to do well in our press daily surveys and do well in patient satisfaction scores we can tell the patient no no problem go ahead become pregnant. We got you. The concern there of course is there, there is risk to pregnancy with severe valve or heart disease. So that might be exposing the patient to more risk than than she might be comfortable with. We try to address the hemodynamics prior to conception and recommend a preconception bioprosthetic aortic valve replacement but as we'll talk about, we can anticipate accelerated valve degeneration and young women who get bioprosthetic valves. If we want to avoid the problem of early degeneration we could choose for durability and recommend a mechanical aortic valve before pregnancy, but as we'll talk about in some depth. There's high fetal risk of maternal from body complications, and there's high fetal risk from the anticoagulation that we need to use for women with mechanical heart valves. We'll get into that in some detail. And increasingly what we are recommending is a preconception Ross procedure but that really is very center dependent it's the right surgery. If you have the right surgeon and the right outcomes. I think there still needs to be more. There's, there's more to be learned about the widespread application of the Ross procedure for young patients with valve of the heart disease in terms of long term outcomes. So what is, what is the right choice for a patient like this. So patients with valvular stenosis are going to increase their gradients over the course of pregnancy and this makes sense if we think about it because we know that both stroke volume and heart rate are going to increase their pregnancy. So in this top left graph, we have a graph of stroke volume here in the y axis from the time of conception, all the way through pregnancy and into the postpartum period. And what you can see is that stroke volume peaks around the end of the first into the second trimester and then plateaus and remains high throughout the rest of pregnancy that curve for heart rate looks similar heart rate goes up during pregnancy as cardiac output is the product of heart rate stroke volume, we can expect cardiac output to go up peak towards the end of the second trimester, and remains considerably elevated throughout the rest of pregnancy. So women with a fixed valvular stenosis are going to have increased gradients across their valve, even with a fixed aortic valve area. So here was one study that looked at systolic pressure gradients in women with aortic stenosis at three time points in pregnancy. We have the mean systolic pressure gradient in the y axis and the time and pregnancy in the x axis and what you can see is that the women who started with a mean gradient solidly in the moderate range in the high 20 millimeters mercury had an increase in their pressure pressure gradients by 30 to 50% over the course of pregnancy and many of these women fulfill the criteria for severe aortic stenosis by their second trimester with mean systolic pressure gradients over 40 millimeters mercury and it remained elevated throughout pregnancy and this may be well tolerated in many women, but in the patient with a vulnerable ventricle or at the edge of decompensation or ventricular uncoupling, this could in theory precipitate heart failure and decompensation over the course of pregnancy. So we know the gradients will go up. And not surprisingly with increasing gradients and cardiac output during pregnancy aortic stenosis we can anticipate will expose both the mother and the fetus to risk of increased complications during pregnancy. And, thankfully, we don't leave it to each individual clinician to figure out which cardiac lesion is the riskiest during pregnancy, but fortunately we do have risk scores to identify maternal characteristics that put women at risk of adverse pregnancy outcomes. The three most commonly used risk scores that we talked about when we work in the field of cardio obstetrics are the Zahara risk score which is the European score published in the European Heart Journal in 2011. CAR PREG which was recently updated to the CAR PREG to this comes out of the group in Toronto, and the most recent iteration was published in 2018, and then the World Health Organization has also put forth a risk classification schema. And CAR PREG and Zahara are similar in their methodology. They each studied a cohort of women with cardiovascular disease to determine which maternal features increase the risk of pregnancy, came up with a set of clinical characteristics and then applied that to a validation cohort to delineate which maternal characteristics predicted risk to the mom and the fetus during pregnancy. The World Health Organization risk score is a little bit of a different schema it's more of an expert consensus where they took a collective of folks who care for women with heart disease and pregnancy and categorized each lesion in a one to four scoring system one being essentially no substantially increased risk of adverse outcomes during pregnancy this would be things like a repaired ASD, VSD or isolated PACs or PVCs, all the way up to risk categorization for which would be considered the lesions which are prohibitive risk for pregnancy this would be things like pulmonary hypertension or severe systolic heart failure with class three or four symptoms. And it's important to notice that all for all three of these scores identify maternal aortic stenosis is a high risk condition so our instincts are correct AS is high risk for pregnancy. But again that's really insufficient to provide adequate counseling for a woman with aortic stenosis who wants to get pregnant because what they're going to want to know is how high is the risk and what are the risks. Importantly, Carpreg and Zahara use these combined outcomes, which combine very important risks like death and stroke and permanent disability with probably less important risks things like maternal arrhythmia and need for new diuretic. And it's very hard to take that combined endpoint and counsel a woman about a risk for pregnancy because some of those outcomes are going to matter much more than others. So we were a little bit in the dark until the row pack registry came along. And the row pack registry for those of you who don't know has been this this this transformational registry out of the European society of cardiology it's the registry on pregnancy and cardiac disease. And it is a prospective global registry of pregnant women with structural or congenital heart disease. And since 2007, they've enrolled over 5700 pregnancies from 138 centers and 60 countries. And by virtue of being a prospective registry, it's been pivotal because it avoids all the problems of recall bias and reporting bias and publication bias that we saw with retrospective or single center studies. The row pack registry is really changed our knowledge base in terms of how we quantify risk in women with heart disease going into pregnancy and in 2016, the row pack registry published their registry on pregnancy and cardiac disease as it relates to aortic stenosis. And this included 96 women with moderate or severe aortic stenosis they published the results in Jack. And the population consisted of 62 women with moderate as and in that group the peak pressure gradient was 48 millimeters mercury but there were 34 years 34 women with severe aortic stenosis prior to pregnancy. And these women actually had quite severe aortic stenosis the peak gradient this cohort was 89 millimeters mercury. The majority were asymptomatic 96% New York Heart Association class one or two, and as a whole they had preserved systolic function. So if we jump to the results, the top line result of the row pack paper on aortic stenosis is that there was no maternal mortality. So what we have is three columns here, the entire cohort on the left those with moderate aortic stenosis in the middle and severe aortic stenosis on the right. On the screen you can see that there was no maternal mortality across any of the cohorts and obviously this was fantastic news. I'd be remiss not to mention that there was substantial maternal morbidity in the cohort. So if you look at maternal cardiac admissions was 21% across the cohort 11% with heart failure and 21% with preterm birth. You can see that the cohort was severe aortic stenosis on the right you can see not surprisingly those numbers are even higher, more than a third had an unplanned cardiac admission, nearly a fifth had heart failure and preterm birth and more than a third of them. So not in not a non morbid condition but thankfully one without mortality. So in the results of the paper the authors submitted a proposed schema in their central illustration that you're all familiar with from Jack. And here's what they recommended is what they consider to be a reasonable pathway for and I concur with this I think it's, it's a reasonable approach, which is, if you have a woman with less than severe aortic stenosis here a maximum velocity of less than four meters per second. She is asymptomatic and she has a normal left and trick or ejection fraction, you can probably counsel her that it's reasonable for her to go ahead and become pregnant without additional preconception testing. If on the other hand you have a woman who comes to you preconception with severe aortic stenosis and maximum velocity more than four meters per second, who is asymptomatic as a normal EF. You're getting an exercise test and a BNP and if those are normal, go ahead and advising that pregnancy is reasonable risk. On the other hand if they're symptomatic or their exercise test is abnormal or their BNP is abnormal, then you would counsel against pregnancy and again this makes sense because severe symptomatic aortic stenosis or severe aortic stenosis with non normal exercise but BNP is an indication for aortic valve replacement and a dictum is that if someone has an indication for intervention prior to pregnancy you should complete that intervention prior to pregnancy. But what happens if a woman comes to you during pregnancy with severe aortic stenosis. If she's asymptomatic with a normal EF, then you just monitor through pregnancy. However, if she's symptomatic, they recommend trying medical therapy although we need to recognize that there's really limited medical therapy available to us for severe symptomatic as there's really just volume management. So that fails moving forward with either balloon aortic valvealplasty or aortic valve replacement or likely in the current era transcatheter aortic valve replacement during pregnancy. We're not going to talk much about AVR during pregnancy but we can discuss it at the end, if folks are interested. So this was a big paper was published in Jack it was important it was the most impactful paper on aortic stenosis and pregnancy that I think have been published and today, and it got a little bit of press. So the European Society of cardiology at their meeting reported this as a breaking clinical trial and many of you probably receive cardiology news carefully shrink wrapped into your mailbox each week and their headline was the ESC says there's zero death in pregnant women with severe aortic stenosis, which was exciting, except we all know that that's not quite true. We know it because of individual case reports this one showing women a woman with severe aortic stenosis who died shortly after delivery. A 24 year old woman with sudden death and the third trimester from severe aortic stenosis and from our own center, a 23 year old woman who died in the postpartum period with severe aortic stenosis. There's certainly no death in that cohort of 90 some odd women with aortic stenosis only 30 of whom had severe aortic stenosis, but I think it would be premature to say that the risk of debt is actually zero. So taking that information. What do we say when we come back to our patient which of these choices would be right for our patient. Well she's asymptomatic preserve the direction fraction, but she is has severe aortic stenosis. I think in our clinic we would get an exercise test in a BNP and if those were normal, we probably would counsel towards pregnancy. And if the exercise test or the BNP were abnormal, then we would probably counsel towards some type of valve restoration prior to pregnancy and in our center I think we would recommend a Ross procedure if she were, if you were anatomically suitable for it. So this is consistent with with what the central illustration of the or what paper had to suggest. So what are the choices let's say we decide that she does need an aortic valve replacement prior to surgery. Well let's talk a little bit about the pros and cons of a bioprosthetic aortic valve replacement prior to surgery. Well, the good news of course is that it's going to restore the hemodynamics prior to pregnancy and will probably be effective at getting her through her pregnancy uneventfully. And of course, the good news is that after the first few months there's no need for long term anti coagulation throughout pregnancy. There's considerable trade offs we know that bioprosthetic valves generate faster and younger patients. There's really conflicting data as to whether pregnancy may accelerate valve degeneration. And if we put a bioprosthetic aortic valve in a young woman, we're likely setting this woman up for many repeats sternotomies over the course of her lifetime and while this may be reduced in the era of valve and valve Tavir. We don't really play out and we don't really know what the long term outcomes of valve and valve Tavir are in young patients. So here's some data showing the bioprosthetic aortic valves to generate faster and young patient. This is a remarkable single center study of over 12,000 patients with bioprosthetic aortic valves stratified by age and implantation. The x-axis is structural valve degeneration, the probability of structural valve degeneration in the x-axis is time. And what you can see is that patients who were under age 60 at the time of implantation highlighted in orange here are much more likely to experience structural valve degeneration and patients older than 60 and certainly than those older than 80. As I said, it's uncertain whether pregnancy accelerates valve degeneration. So while a preconception bioprosthetic valve may result in good pregnancy outcomes, repeated sternotomies are likely. Well, if we want to avoid that cycle of faster generation of bioprosthetic valves, I suppose we could recommend a mechanical prosthetic heart valve prior to pregnancy, but of course, mechanical heart valves put both the mother and the fetus at risk. So on the left, we have the maternal rests of mechanical heart valves in pregnancy, namely valve thrombosis and ball of complications and hemorrhage. And on the right, we have the fetal rests of miscarriage, fetal demise and teratotranic effects of anticoagulants. And when we talked about this, we need to remember that pregnancy is a hypercoagulable state with an increased risk of thrombosis and poor anticoagulation options. And even though women may have a normal PT, PTT and INR, they have anything but normal coagulation profile. So here's a table just showing sometimes the very dramatic up to 10-fold changes in coagulation factor activity during pregnancy, and we know that this overall translates into a two-fold increase in coagulation activity, which is the cause behind the five-fold increase of DVT and PE that we're all familiar with during pregnancy. So mechanical heart valves have an accelerated rate of complications during pregnancy. So prosthetic heart valves put the mother and fetus to risk. And if we want to know exactly how much risk, again, we're fortunate to be able to turn once again to that Ropak registry. In 2015 in circulation, the Ropak authors published their data on the outcomes of women with mechanical heart valves in pregnancy. This paper doesn't stratify by different anticoagulation strategies very well. It lumps together all the women with mechanical heart valves who became pregnant. And remember, this is in a contemporary cohort. The outcomes were not very good. 23% had a cardiac hospitalization, 15% had a major hemorrhage, 5% had valve thrombosis and strobe in one and a half percent. The fetal outcomes were not better. 16% miscarriage and 3% field demise after 24 weeks. The maternal birth rate of 18% and a congenital abnormality rate of 5% probably due to a combination of teratogenic effects of anticoagulation combined with the heritability of maternal cardiovascular conditions. So high rates of maternal, high rates of field complications. But very importantly, the maternal mortality over the cohort was 1.4% and the major event rate to the mother or the fetus was 42%. And remember, this is over the course of just nine months. This is not the lifetime risk. This is not the 10 year risk. This is just the risk of adverse events over the course of a single pregnancy mortality at 1.4%. So because of this, the conventional wisdom is that when possible it makes sense to avoid mechanical heart valves and women planning pregnancy. Okay. So if we help women with mechanical heart valves who are committed to pregnancy, or alternatively, what do we say to an already pregnant patient who wants to find the safest anticoagulation strategy for her and her baby. We start with a primary dictum which is uninterrupted effective anticoagulation is required throughout the entire pregnancy for all women with mechanical heart valves. We talk about anticoagulation for mechanical heart valves of course we all use warfarin which is an effective anticoagulant but unfortunately is toxic for the developing fetus. Just as a brief reminder about warfarin embryopathy warfarin embryopathy occurs primarily when the fetus is exposed between week six and 12 of the 10% risk of deformity 20% risk of loss. It's characterized by mid face hyperplasia short limbs, you have an anticoagulated fetus and that's true regardless of what stage in pregnancy the woman is taking warfarin. And there are variable reporting as to whether there's intellectual impairment in babies born with warfarin embryopathy. So, what are our strategies, well, we want to mitigate the risk of warfarin on the fetus while still providing adequate anticoagulation to the mom. So fundamentally, there's three strategies we can do. If our primary objective is to minimize the possibility of toxic effects of anticoagulants at the fetus, we could use heparin for all three trimesters. So, this slide breaks it up into first trimester third second trimester and third trimester heparin all three trimesters heparin and low molecular weight heparin are large molecules they don't cross the placenta they don't touch the fetus. So, there's no risk to defeat us, but it's not as good an anticoagulant from mechanical heart valves. If we want to strike some sort of balance by minimizing the risk of warfarin embryopathy but anti coagulating the mother well we could use heparin in the first trimester and warfarin in the second and third trimester. And if we want the most effective anti coagulation strategy for mom for all three trimesters, we could use warfarin throughout. Okay. In February we always switch to IV heparin because you don't want to deliver a woman anti coagulated with warfarin. So, we'll start just by talking briefly about this pivotal meta analysis that came out 20 years ago now in the archives of internal medicine. This was a meta analysis published by Chan at all of 28 papers published between 1966 and 1997 looking at 1234 pregnancies of women with mechanical heart valves and because of the era that the papers were published half of these women had had two thousand cage valves, two thirds or mitral valves, and overall in that cohort the maternal mortality was 3%. So what were their results here. So what we have is the three anti coagulation strategies and the rose heparin on top. Heparin in the first trimester transitioning to warfarin in the middle and warfarin for all three trimesters on the bottom. And then we have three columns of maternal death, death, thromboembolism and congenital anomalies. What we found is that when you look at maternal outcomes. They're worse in women who take heparin all three trimesters they're intermediate in women who take heparin for the first trimester and warfarin for the next two. And they're best in women who take warfarin for all three trimesters. And unfortunately and not surprisingly you see the exact opposite trend as it relates to fetal outcomes, where they're best in women who take heparin all three trimesters and worse than women who take warfarin for all three trimesters. So there's this balance that you're trying to strike where any benefit you get maternal outcomes is offset by fetal risk. So the take home risks the take home messages from that 2000 meta analysis is that pregnancy no matter how you do it is high risk for women with mechanical heart valves. Fractionated heparin throughout pregnancy with that 7% maternal mortality rate at unacceptable maternal outcomes we just should not be using three trimesters by the heparin or subcutaneous unfractionated heparin in pregnancy and warfarin for all three trimesters associated with poor fetal outcomes. It's a really impactful paper but we need to recognize that a lot has changed since the 1960s we no longer use ball and cage valves which are highly thrombogenic we've moved to these bi-lethal disc valves. We no longer use IV heparin routinely in pregnancy we've switched to low molecular weight heparin with its more stable half life and better, better pharmacokinetics. Unfortunately not everything has gotten better since the 1960s I was just lucky enough to go see Paul McCartney in concert a couple weeks ago if he does come down to you guys I highly recommend it. That was music in the 60s and Bieber is what we got now. With this in mind a couple years ago one of my fellows and I who's now faculty with us Zach Steinberg decided to update this paper with the goal of comparing contemporary anti-coagulation strategies for mechanical heart valves for women in pregnancy and we wanted to describe maternal and fetal outcomes that were of interest to the patients and doctors. So we didn't repeat the information that was gleaned from the 2000 paper we decided to be more exclusionary in the papers that we included we decided we were going to include only papers that reported unambiguous outcomes and women with mechanical heart valves who were treated with either PKA warfarin or low molecular weight heparin or unfractionated heparin and we were going to exclude studies that included a high number of ball and cage valves, high number of right sided valves small studies previously published and those that use fixed those heparins. This is what we came up with our original search yielded 825 papers and we were able to immediately exclude nearly 600 of them. Another hundred were review articles are not terribly useful. We were left with 155 papers which seemed like a lot but when we went through them. A large number were small case series had ball and cage vows were did not report carefully on outcomes or were otherwise not not usable and we were left with 18 papers over describing 800 pregnancies which is still a pretty large number of pregnancies but it just shows how hard it is to find high quality data in this field. What we did is we compared four strategies of anti-coagulation during pregnancy. We looked at warfarin throughout and then we stratified that among women who require more or less than five milligrams of orphan per day to maintain a therapeutic ion or and I'll talk about that more in a moment. We looked at those adjusted low molecular weight heparin throughout pregnancy. We looked at low molecular weight heparin transitioning to warfarin. And then we looked at unfractionated heparin transitioning to warfarin for the second and third trimesters. So we wanted to find out which of these four strategies gave the best results for the mom and for the fetus. We decided to look at outcomes that we thought would be unambiguously important to both the women and the doctors. We looked at maternal death, we looked at major embolic complications and we looked at prosthetic valve failure requiring reintervention. And then for fetal outcomes we looked at spontaneous abortion that is lost before 20 weeks and fetal death that is lost after 20 weeks. And then we looked at congenital defects and embryography. So this was our central illustration. And what we found is that the risks for the mother and the fetus were high for all the strategies. So first we'll look at the risk for the mother. And these are a graphical depiction of the percentage risk of a major event to the mom with each of the strategies warfarin and low dose warfarin, both had about a 5% risk of a major complication. And all the strategies that used heparin had a nearly three fold increase in maternal risk. Of course, when you look at the risk to the fetus it goes in the exact opposite direction. And K antagonists had a 35% risk for the high dose cohort where it was cut in half and women were able to take low dose warfarin low molecular weight heparin and low molecular weight heparin with a VK a had lower risk and unfractionated heparin to a VK still had very high risk and that may reflect the parts of the world in which unfractionated heparin were were still used. So with the combined outcomes to either the mother or the fetus, the risks are high no matter what strategy you use high dose warfarin has a more than 40% risk to the mother the fetus. The lowest cumulative risk was for the women who were lucky enough to maintain a therapeutic ion are on low dose warfarin, and those that used heparin had high cardiovascular risks. So, there was no low risk strategy and everything was a trade off warfarin being safest for the mom heparin being safest for the fetus. Right after we published our paper group out of Europe published a very similar meta analysis in the European heart journal and we were relieved to show that it showed more or less the similar results. We came up with 46 studies with nearly 2000 women they were a bit more inclusive in their inclusion criteria with 2500 pregnancies. When we look at warfarin versus sequential treatment versus heparin for all three trimesters, you see this familiar trend warfarin safest for the mom. Lomeloid heparin is riskiest for the mom and fetal loss is highest with warfarin and lowest with Lomeloid heparin. They had a nice figure that was able to stratify fetal risk according to low dose warfarin versus high dose warfarin. And what they found is that low dose warfarin in green was associated with a higher probability of having a live birth and lower pathogenic effects. So this builds on evidence base that low dose warfarin is safer than high dose warfarin as it relates to fetal outcomes. So one question is, why is Lomeloid heparin so ineffective I mean dose adjusted Lomeloid heparin is a pretty effective anti coagulation option. And Uriel Kain's group has published extensively on this they are strong advocates of low molecular weight heparin throughout pregnancy. And what they argue is that we are targeting too low an anti 10 a level that if we target a peak anti 10 a level of 0.8 to 1.2 patients can at least spend a large portion of their day sub therapeutic in terms of anti coagulation. So this bar chart here in the middle of the screen, the y axis is the percentage of your time you are sub therapeutic that your trough level will be sub therapeutic depending on what your peak anti 10 a level is. And what you can see is that if your peak anti 10 a level is less than one, there's a very high probability that your trough will be sub therapeutic. Even when you're in the range that we recommend, half the time your trough level will be sub therapeutic and it's not until your peak 10 a is above 1.2. It's only therapeutic throughout the day. So increasingly and at our centers, we will not just use peak levels will use trough levels and we'll try to Goldilocks dial in that just perfect level of low molecular weight heparin. I just need to reemphasize here that there's absolutely no role for weight based fixed those low molecular weight heparin in pregnancy for mechanical heart valves it's absolutely contra indicated you have to meticulously follow anti 10 a levels. So, what's the data on low risk warfarin being better than high risk warfarin, low dose warfarin being better than high dose warfarin as it relates to fetal outcomes well this was a highly cited paper from the tally at all from now 23 years ago they looked at 58 pregnancies stratified by whether women were able to maintain a therapeutic anti 10 a on low dose or high dose warfarin, these women took warfarin through all three trimesters for the women who were able to maintain a therapeutic high dose warfarin in the UGR in less than five milligrams per day 28 of 33 pregnancies resulted in a healthy fetus and five out of 33 had complications but those complications are relatively mild things like small for gestational age or I UGR. You can trace that with the pregnancies and women, we took high dose warfarin throughout pregnancy. What you saw is that only three of 25 ended up with healthy babies on delivery and 22 out of 25 had complications including embryo apathy spontaneous abortion and still birth. So low dose warfarin really does seem to be safer than high dose warfarin as it relates to fetal outcomes. However, once again, case reports can prove that the exception to the rule. So here are four carefully reported cases of women who took low dose warfarin throughout pregnancy and had babies born with warfarin embryo apathy. So again, as we counsel women we need to remember that none of these rules are absolute that the risk of warfarin embryo apathy is lowest with low dose warfarin but it's certainly not zero. So in 2020, we were tasked with updating the valvular heart disease guidelines with trying to take all of this data and update the 2014 valvular heart disease guidelines to reflect what the new data showed in terms of contemporary practice. In terms of what we should be doing for anticoagulation for valvular heart disease in pregnancy and thank you to great mentors for Catherine Otto for including me in this. So we wanted to update the guidelines to reflect contemporary science and eliminate reference to antiquated therapies like three trimesters of unfractionated heparin. And we also wanted to acknowledge that there's no perfect options and that we really needed to value patients priorities. So here's some of the flow charts I just wanted to share from the valvular heart disease guidelines. For women with mechanical heart valves, we recommended a class one recommendation for saying that they should receive therapeutic anticoagulation with frequent monitoring throughout pregnancy I think that's fairly non controversial. And then we said that we should ask whether women really can maintain that therapeutic anticoagulation and frequent monitoring throughout pregnancy because some women cannot for whatever reason, either lack of access to healthcare resources or social determinants of health. Some women can't do it in that case, you should counsel against pregnancy in that population. Even for women who can maintain therapeutic anticoagulation frequent monitoring. We put in a class one recommendation that you should counsel the woman that there is no anticoagulation strategy that is safe for the mother and the fetus and you should really engage and share decision making as to whether pregnancy is consistent with the women's goal and our appetite for risk during pregnancy. So let's look into what to do if the woman does become pregnant or you encounter a woman with a mechanical heart valve who is pregnancy, who is pregnant excuse me. So if their dose is less than five milligrams per day. That's this flow chart on the left that they're able to take less than five milligrams of warfarin per day, the preferred anticoagulation strategy is to continue warfarin for all three trimesters. In fact, we recognize especially here in the United States the, the risk tolerance for birth defects is low, lower than it is in Europe and many other parts of the world. So after a careful discussion we allowed for a to be recommendation for those who does the bone molecular weight heparin for the first trimester, switching to warfarin for the second or third trimester but did not provide an option for low molecular weight heparin for all three trimesters. Then we said okay if they take more than five milligrams of warfarin per day, do they have access to dose adjusted low molecular weight heparin with monitoring of 10 a level that's true almost everywhere in the United States but we recognize that people in low resource countries, maybe accessing this document to. So if they can take low molecular weight heparin, we recommended low molecular weight heparin for the first trimester, and then switching to warfarin for the second and third trimesters is a to a second trimester, but once again recognizing that there is some fetal risk associated there, we did allow for women who wish to minimize fetal risk to take low molecular weight heparin for all three trimesters. And then for women who don't have access to low molecular weight heparin, you can use continuous IV heparin for the first trimester and then warfarin for the second and third trimester, but I think this is going to be uncommon in the United States. So you can stop warfarin and switch to heparin and then you hold the heparin prior to delivering. So those were our recommendations and when I think back and look at the trade offs inherent to anti coagulation strategies and pregnancy it reminds me of a dictum I was taught in 11th grade by my environmental science teacher, it's that there's no such thing as a free lunch I think this comes out of a story that I was told that, you know, back in the late 1800s they would build saloons near construction sites and they would serve free lunch but the free lunch was high in salt and it would make people thirsty and people would buy beer. And then somehow the free lunch would end up costing more than if you just paid for and brought your own lunch from home. And that's something for nothing and that's true not just in free lunches and pregnancy and heart valves but it's true in multiple aspects of life, the thinner your phone the faster the battery is going to wear out by the end of the day. What I want to know is that when it comes to anti coagulation strategies for women with mechanical heart valves and pregnancy, everything is a trade off you cannot simultaneously minimize the risk to the mom, and minimize the risk to the fetus, everything comes at a cost. So with that in mind, what are we going to tell our patient. And here I want to pivot a little bit for the last few minutes to talk, not just about what we tell our patients. But about how we tell our patients things impacts the decisions that they make and what they hear. And now let's pretend our patient comes to us she's got a mechanical heart valve and she's asking. Hey doc, you know what do you think what's the risk of me having a successful pregnancy with a mechanical heart valve. Well, here's a list of three true statements, you could make to your patient statement one pregnancy for you would be high risk. That's indisputable and pretty non controversial statement statement to that's also true. The most likely outcome of the pregnancy is that you and your baby will both be fine. That's statistically true no matter how bad we are choosing an anti coagulation strategy. The odds are that both mom and fetus will be fine. In summary, you're more than 100 times more likely to die during your pregnancy than a woman without a heart condition. So these are three true statements, these are all accurate. But you can imagine that each of those statements may motivate the woman to make a different choice. Patients don't always hear what we mean to say. And I think we all know this intuitively but the people that really know this literature or malpractice lawyers. So if you read in the journals of medical legal folks, they spent a lot of time thinking about how patients perceive risk at the time of informed consent. And what they found is that there's an enormous amount of inter subject variability people hear very different things to the same words, and the words we use are often very very imprecise so let's come up with the word rare, for example. If we are talking to a patient about a procedure and medication or risk of pregnancy and we say that the chance of an adverse outcome is rare. What they interpret varies by orders of magnitude, and those differences depend heavily on the overall health status, age and demographics of the person you're talking to. On average if you tell someone something is rare, the risk of a complication is rare. On average, they think it's about one in a thousand. If you tell that to a healthy 24 year old, their assessment of what rare means is much less than one in 1000. However, if you tell it to a 75 year old with cancer, their estimate of a rare complication is much higher. So words like rare imprecise and we may have no idea what the patient interprets the word rare to mean. So, even worse than using imprecise words is probably using relative risks. This applies both to talking about your chance of dying is 100 times sold higher with pregnancy with a mechanical heart valve as it is with someone with normal heart disease as it does to talking about lightning strike so here is a map of the probability of dying by a lightning strike anywhere in the United States and you know, look, I'm thrilled to be living up here in the top left corner where my risk of dying by a lightning strike is just about zero. And my risk of dying by a lightning strike is more than 1000 times lower than where I live in Florida so I'm making a great choice, but I think it would be a little insane to say that folks should choose where they live based on the relative risk of dying from being hit by lightning, because the relative risk is relatively irrelevant in this situation. So both patients and doctors to be perfectly honest can be misled by relative risks, we're much better off talking about absolute risks. Personally, composite endpoints are really not another really important problem when we talked about risk so when I talked about car Craig and Zahara at the beginning they talked about the risk of an adverse pregnancy outcome. And that included maternal death and it also included maternal arrhythmia for which a woman might need to take a little bit of beta blocker it's difficult to counsel a woman about the relative risks of each of those outcomes because those risks are so meaningfully different in terms of magnitude, as it affects a woman's life a woman very well may be willing to have an unplanned hospitalization or an arrhythmia but unwilling to have a stroke or maternal death or need to go for emergent cardiac surgery during pregnancy. So these composite endpoints are dangerous to when they come to counseling women in terms of outcomes and not just women with pregnancy but patients overall patients don't be risks is equivalent, and we shouldn't make too many assumptions about which risks are most important to women. Some women may be willing to accept much higher risk some women may be willing to accept higher risk of adverse fetal outcomes to avoid risk to the mom and some women may have a totally different set of priorities so we should really make sure to ask about the patients priorities. So what can we do to help patients understand risk better well, we should avoid using these date descriptive terms like low risk but you simultaneously need to be careful about avoiding a giant data dump, where you give them every piece of data that overwhelms them and prevents them from being able to digest and make a sensible decision. When possible, you should use absolute numbers of risk and avoid talking in relative risks. And then finally, we should think a little bit about how we frame the risk. So we talked about the difference between positive framing, which encourages someone to make a risky choice and loss framing which discourages someone to make a risky choice. An example of positive framing is the chance that your heart valve would be fine during pregnancy is 93%. An example of risk framing would be say you have a 7% chance of having your valve fail during pregnancy and needing an emergency valve surgery. Not surprisingly, a helpful way to do it is to frame both say there was a 93% chance your valve would be fine and a 7% chance your valve would fail. It's a couple extra words but it turns out that when you study this, it actually does impact how patients make decisions. You can consider visual aids like this here, 6 out of 100 people in a situation like yours will have an outcome such as death or emergency surgery, etc. So these visual aids can help people, not just folks with low medical literacy, but folks in general really start to internalize what it means when we talk about the absolute risks of outcomes. So in summary briefly, I think if I had to leave us with three take home points I'd say the guidelines direct us towards the safest strategy, but pregnancy and women with balveolar heart disease can be risky and pregnancy with mechanical valves has no best option. There ain't no such thing as a free lunch. And we need to be thoughtful about how we communicate risk and how we listen to the patient's priorities. So, honestly, thank you. It's an honor to be invited down here and be able to get grand rounds to you. It's a topic I share a lot about so I thank you all for joining in for your for your attention today. Thanks so much.